Project Summary/Abstract Metabolic reprogramming is a hallmark of malignancy and a source of therapeutic targets. Progress in translating reprogrammed activities into new therapies is limited by the fact that the vast majority of knowledge in tumor metabolism is derived from studies in cultured cells with unknown relevance to disease biology. We developed methods to assess metabolic flux directly in tumors from human subjects and mice, thereby eliminating artifacts of culture. Our approach integrates multi-parametric imaging of the tumor with intra- operative infusions of isotope labeled nutrients like 13C-glucose. We use information from pre-surgical imaging to guide tissue sampling, so that we can assess the effects of relevant biological features (glucose uptake, perfusion, tissue density) on tumor metabolism. After surgery, we perform a fragment-by-fragment analysis of tumor and adjacent lung to measure fluxes and examine their relationship to histology, genetics and gene expression. Our published work in human non-small cell lung cancer (NSCLC) demonstrated that a) contrary to long-held expectations, these tumors oxidize glucose in excess compared to adjacent lung; b) tumors oxidize other fuels in addition glucose, and regional fuel choice is predicted by pre-surgical imaging; and c) extensive metabolic heterogeneity exists among human lung tumors and even within distinct regions of the same tumor. As far as we know, our multidisciplinary approach integrating clinical imaging with metabolic flux analysis, quantitative histopathology and molecular features is unique. Here we propose to expand our program in human NSCLC metabolism to address emerging, pressing questions over the next several years. We are establishing novel computational methods to better report altered fluxes throughout the complex metabolic networks of human NSCLC. We are establishing a series of NSCLC xenografts from patients recruited to the study, providing us with a biological test bed for hypotheses stimulated by observations made in the clinical studies. In patients and mice, we will examine the evolution of metabolic phenotypes during cancer progression, including by sampling metabolic flux before and after conventional and targeted therapies. We have identified a number of candidate fuels and are now infusing patients with a series of 13C and 15N-labeled nutrients to test which ones are consumed by tumors and how their metabolism is regulated in vivo. Finally, we are establishing methods to disentangle the metabolic contributions of distinct cell types comprising the tumor microenvironment in patients and mice. We believe that understanding metabolic crosstalk between cancer and stromal cells in the intact tumor microenvironment is one of the most daunting technical challenges in the field, but also the best opportunity to make fundamentally new discoveries. Altogether, these efforts will generate a unique view of NSCLC metabolism with an unprecedented level of detail, biological accuracy and relevance to human disease. They have the potential to establish new paradigms in metabolic regulation and tumor heterogeneity and to predict which patients will respond to metabolic therapies.